Semantic-Aware Video Text Detection
Wei Feng1,2; Fei Yin1,2; Xu-Yao Zhang1,2; Cheng-Lin Liu1,2
2021-06
会议名称Internaltional Conference on Computer Vision and Pattern Recogintion
会议日期2021.6.19
会议地点线上会议
摘要

Most existing video text detection methods track texts with appearance features, which are easily influenced by the change of perspective and illumination. Compared with appearance features, semantic features are more robust cues for matching text instances. In this paper, we propose an end-to-end trainable video text detector that tracks texts based on semantic features. First, we introduce a new character center segmentation branch to extract semantic features, which encode the category and position of characters. Then we propose a novel appearance-semantic-geometry descriptor to track text instances, in which semantic features can improve the robustness against appearance changes. To overcome the lack of character-level annotations, we propose a novel weakly-supervised character center detection module, which \textbf{only uses word-level annotated real images} to generate character-level labels. The proposed method achieves state-of-the-art performance on three video text benchmarks ICDAR 2013 Video, Minetto and RT-1K, and two Chinese scene text benchmarks CASIA10K and MSRA-TD500.

收录类别EI
七大方向——子方向分类文字识别与文档分析
文献类型会议论文
条目标识符http://ir.ia.ac.cn/handle/173211/45047
专题模式识别国家重点实验室_模式分析与学习
作者单位1.National Laboratory of Pattern Recognition, Institutaion of Automation of Chinese Academy of Sciences
2.School of Artificial Intelligence, University of Chinese Academy of Sciences
3.CAS Center for Excellence of Brain Science and Intelligence Technology
推荐引用方式
GB/T 7714
Wei Feng,Fei Yin,Xu-Yao Zhang,et al. Semantic-Aware Video Text Detection[C],2021.
条目包含的文件 下载所有文件
文件名称/大小 文献类型 版本类型 开放类型 使用许可
03057.pdf(2299KB)会议论文 开放获取CC BY-NC-SA浏览 下载
个性服务
推荐该条目
保存到收藏夹
查看访问统计
导出为Endnote文件
谷歌学术
谷歌学术中相似的文章
[Wei Feng]的文章
[Fei Yin]的文章
[Xu-Yao Zhang]的文章
百度学术
百度学术中相似的文章
[Wei Feng]的文章
[Fei Yin]的文章
[Xu-Yao Zhang]的文章
必应学术
必应学术中相似的文章
[Wei Feng]的文章
[Fei Yin]的文章
[Xu-Yao Zhang]的文章
相关权益政策
暂无数据
收藏/分享
文件名: 03057.pdf
格式: Adobe PDF
所有评论 (0)
暂无评论
 

除非特别说明,本系统中所有内容都受版权保护,并保留所有权利。